Automated standardization of images of Drosophila embryos

Modeling expression patterns of Drosophila, in space and time, plays a critical role to understand the development of multicellular organisms. In confocal microscopy, to produce precise quantitative data it is frequently necessary to process and analyze large amounts of digital images. Automatic preprocessing is a crucial step in this scenario, essential to standardize significant features such as orientation, size, position, direction, lighting condition and texture of embryo images. Even though a lot of efforts have been made, a robust embryo standardization strategy is still needed. In this paper, we propose the method Embrystandar. It is designed to remove background artifacts and standardize the direction and orientation of a Drosophila embryo through a sequence of automatic operations. To test its potential for large-scale image processing, Embrystandar was applied in different databases. It showed to be robust and precise, reaching more than 90% success rate.